When we'll see elinks for MSX?

Page 3/7
1 | 2 | | 4 | 5 | 6 | 7

By Pencioner

Paladin (711)

Pencioner's picture

11-11-2018, 22:00

Grauw wrote:

At least, I don’t know how difficult it is to implement TLS on a Z80 (not familiar with the algorithm).

it is an overload for Z80, even just RSA handshake is already barely possible (even with Turbo R i guess). And after handshare you should encrypt everything with strong encryption like AES 128/256 bit etc. Impossible...

By Grauw

Enlighted (7548)

Grauw's picture

12-11-2018, 00:00

As I understand it from Wikipedia RSA is assymmetric (slow) and AES is symmetric (fast). RSA for the handshake, AES for the data. A byte oriented AES algorithm implementation seems to be this:

https://storage.googleapis.com/google-code-archive-downloads...

Well, not impossible, I guess… I see mostly table lookups and xor operations. Would be fun to implement AES and test the performance on some test vectors. Maybe on par with gunzip?

By AxelStone

Prophet (2641)

AxelStone's picture

13-11-2018, 16:58

Quite interesting these last replies. I really think that basic HTML with basic CSS could generate webs able to be consumed by MSX without any additional hardware. In fact adding hardware is not a good solution, since you've already added GR8Net that is quite expensive, so the idea is to use directly MSX + GR8Net. Anyway the problem still remains: only specific webs could be consumed in MSX. This is why I thought in a text based browser, to avoid complexity.

Perhaps a solution should be to use a custom proxy handled online that translates any web into basic html, so our MSX should point to that proxy?

By Grauw

Enlighted (7548)

Grauw's picture

13-11-2018, 19:12

Maybe such a proxy already exists? At least some browsers have a similar "reading mode" where they can remove distracting parts of sites leaving just the text content, perhaps a website with similar functionality exists as well...

By edoz

Prophet (2097)

edoz's picture

14-11-2018, 09:08

i never found something like a proxy or program that can translate webpages to a simple HTML text output. You can use Lynx to save a website to text output but this does not contain any hyperlinks. I think that Symzilla can support simple HTML but that is very rare on the web nowdays. (The help browser in SymbOS is already HTML based)

By skumlerud

Resident (34)

skumlerud's picture

14-11-2018, 09:22

Considering the way the internet is heading we will soon need a local "retronet" to get our computers online at all.

By edoz

Prophet (2097)

edoz's picture

14-11-2018, 09:39

skumlerud wrote:

Considering the way the internet is heading we will soon need a local "retronet" to get our computers online at all.

I agree with that. That is also the idea in Symzilla. Prodatron use a compiled DOX format which is optimized for browsing. Next to that it could support basic HTML (like in the helpbrowser of SymbOS, which is HTML based) in Symzilla for "normal" internet access. And yes, a proxy would be cool to convert pages to basic HTML.. but that is not easy.

By Grauw

Enlighted (7548)

Grauw's picture

14-11-2018, 11:28

Just saying that I agree it seems better if that "retronet" / proxy was HTML (/ CSS) based rather than a custom format. So that it's accessible in the regular web browser as well, also it makes authoring, testing and deployment easier and cheaper because we can use existing tools, standards, server frameworks and hosting configurations.

By edoz

Prophet (2097)

edoz's picture

14-11-2018, 13:17

I was looking for tools in the past than can convert nowdays sites to basic HTML output. But i was not able to find something. If we can do pure HTML it would be the best of course.

By Grauw

Enlighted (7548)

Grauw's picture

14-11-2018, 16:08

Should not be any more difficult than converting to whichever other format (like DOX), right?

The first hit on my first google search query for "html simplifier":

https://www.htmlwasher.com/

Looks like it does a fine job. According to the explanation at the bottom it’s pretty simple too.

This type of filter could even be built-in on an MSX browser; as the parser processes the input stream it can create the document tree on the fly with only the tags it recognises, it does not need to store the entire file in memory.

And because we can’t show all content at once anyway this can also be done lazily, since it only needs to process a screen’s worth of data to render it. So it’d be pretty responsive. This is also what modern PC browsers do in order to decrease the time between loading and presenting it to the user, and the HTML and CSS standards are designed to support this (e.g. CSS selectors never look forward).

Page 3/7
1 | 2 | | 4 | 5 | 6 | 7
My MSX profile